← Back to Blog

June 19, 2014

Exploring the Variation in R&D-intensity Across Canadian Industries

Hi there – I’m the DEEP Centre’s data analyst and as part of my work I am have been analyzing trends related to business expenditure on research and development (BERD) by industry. As Dan and Warren have often written, this measure of investment is very weak in Canada relative to our OECD peers.

I thought I’d dig into the numbers to see which industries are particular big or small investors in BERD. To measure the relative research-intensity of the industries, the ratios of total sectoral BERD and output (GDP) were calculated. The average research-intensity of industries over the past 17 years is compared here and the results are represented in the charts below.

Here’s what we see:

The IT and Telecom industry has remained the most research-intensive sector in Canada with, on average, 9.65 cents spent on R&D per 1 dollar of GDP  since 1994. This is more than eight times higher than the economy’s average (1.19), and about twice as high as the ratio for manufacturing, the second most research-intensive sector.

While the ratio for manufacturing is weak, it’s the dismal less than 1% share of investment in mining and oil and gas that is most shocking . Given ongoing debates on the environmental sustainability of these industries, and the significant environmental impacts, one might expect to see far more investment in innovation and research than these numbers show.


Moving on, it’s interesting to think about why the BERD/GDP ratio (or the R&D intensity) varies so much between industries. If one believes that a sector’s R&D intensity directly reflects the market demand for new technology in that sector, than sectors will see higher ratios when a high ROI on R&D expenditures can be expected. In fast moving tech/telecom sectors, this ROI is apparent, and somewhat the case in manufacturing where technology-intensive sub-sectors like semiconductor and electronic components have very high ratios, as compared to weak ones in textiles, food (which should be higher given processing advances and the growing demand for novel products) and furniture.

However, there are some surprising cases. We would expect the agricultural sector to be technology-intensive, given significant research in the field related to yields and new product strains. However the average R&D intensity in this sector is three times less than the economy’s average. A similar situation can be seen in the construction sector. And as noted, the energy sector (Mining, Oil and Gas) also demonstrates a surprisingly low research-intensity measure. This latter sector is perhaps the most likely to see government regulation introduced as a means of increasing the level of innovation in the sector.

Finally, it`s only fair to note that the data used for this has its limitations. Depending on the firm and its accounting practices, and its supply chain, research expenses can be incorporated in a variety of means that may or may not be distilled as R&D. An alternative method to measuring R&D intensity of a sector would take into account the inter-sector flow of products by using input/output (IO) tables (if you’re read this far let me know and perhaps I’ll do this in the future!)

All this said, there are real questions as to how BERD can be increased in specific sectors and, given that we already heavily subsidize research, what tools the government has to do this. In some this can rely on regulation regarding health or the environment, in others it might simply be help in recognizing that without increased investment, the competitiveness and sustainability of firms will be on the decline.


Interested in becoming a DEEP Centre member?

The Centre for Digital Entrepreneurship and Economic Performance turns research findings into practical insights and tools for stimulating innovation, entrepreneurship and job creation.